UTYPEs and UFIs

[Francois Bonnarel]

My answers and remarks to the questions adressed by Jonathan in his nice document.
 a) Value fields:  Values are not so often simple for example in STC which char
is using Value is either Value1 or value2 or value3 value1 will be a double but
value2 is the gathering of value2.C1 and Value2.C2, etc ... So what will be adressed by 
the upperlevel Coordinate where value is to be found???

b) combined datamodel: here I think the first use case is the combination of Characterization
and STC for example as in  Coverage.location where coverage.location.coord is actually
a stc:AstroCoords. What does that mean ? that location.coord gathers SEVERAL stc:Coordinate
(usefull in the general case of intertwinned axes ) and of a pointer to the coordsytem.
And that each of these stc:Coordinate can have a Name, Value, Error, Resolution, etc ...
WE cannot avoid to get so three additional level. It's not a problem of complexity but of
lenght. I prefer to integrate the new namespace in the string to show where the new model is
coming. to shorten the whole we can start the char utype where it's non ambiguous
sp that "char:location.coord.stc:Coordinate.error2" will be a valid utype. A software
aware of Char will know that location is a part of CharacterizationAxis.coverage
unambigouisly 

c) UFI : They are a dynamical building from the utypes by fixing some other attribute
in the model to a given value.
If the bracket is considered as bad (because it's too xpath oriented) can we see ufi  as the 
result of a pseudo sql query like this:
UFI=(SELECT COLUMN FROM DATAMODEL WHERE UTYPE="stc:AstroCoords.Coordiante.Value2" and
CONSTRAINT="stc:AstroCoords.coordsystem_id" and VALUE="TT_ICRS"
?????

d ) Utypes in VOTABLE. we can shorten in GROUPS if we adopt my proposal in point b
(start the char utype where it's non ambiguous)




FRançois

=====================================================================
Francois   Bonnarel               Observatoire Astronomique de Strasbourg
CDS (Centre de donnees          11, rue de l'Universite
astronomiques de Strasbourg)    F--67000 Strasbourg (France)

Tel: +33-(0)3 90 24 24 11       WWW: http://cdsweb.u-strasbg.fr/people/fb.html
Fax: +33-(0)3 90 24 24 25       E-mail: bonnarel at astro.u-strasbg.fr
---------------------------------------------------------------------